Osteoarthritis (OA) of the hip and knee are prevalent disorders which result in significant disability and societal cost. The causes of hip and knee OA are multifactorial. That heredity is one of the factors influencing OA susceptibility and progression is not in doubt, even though it may not be the principal factor. The long term goal of this project is to identify genes which account for the heritable component of risk for hip and knee OA and to understand the precise mechanisms by which these genes exert their effect. Aim 1 of this project will determine the risk to relatives for OA by surveying 1700 patients who will have undergone joint anthroplasty for primary/idiopathic OA at University Hospitals of Cleveland between 1994 and 1999. A subset of these patients (150 with knee OA and 150 with hip OA), who have living, clinically and radiographically evaluated, affected siblings will then be used to determine non-hereditary and hereditary contributions to this relative risk. This subset of patients and their previously evaluated affected relatives will be surveyed in detail, have their radiographs reviewed, and have blood collected for DNA extraction. Test of aggregation will be used to distinguish between non-hereditary and hereditary contributions to OA relative risk of each of these two 150 affected relative pair cohorts (knee and hip). A major advantage of this approach is that it is statistically robust and not prohibitive in cost. This is because all affected relative pairs a priori have already been clinically and radiographically evaluated for OA. Additionally, cartilage will be available from half the affected pairs probands since they will be ascertained prospectively during the first two years of this project. Aim 2 of this project, occurring during years three and four, will identify genes which contribute a moderate increase in relative risk for OA using high throughput genotyping to look for regions of excess allele sharing in the affected relative pairs. These regions will be identified by testing candidate genes and by performing a genome wide search at 10 cM resolution. Cohort sizes of 150 affected relative pairs per arthroplasty type (knee and hip) have the power to detect genes contributing a 3-fold increase in risk to relatives. Risk associated mutations/polymorphisms will be sought in the linked candidate genes. Strongly linked regions not presently containing candidate genes will be a focus for future studies. Identifying genetic variants which increase the risk of developing common hip or knee OA, and understanding the processes by which this risk is increased, will suggest new approaches for preventing and treating OA.